Maximising Uptime on a single server

I am thinking of the best method for maximising uptime on a single server (running Win2K).

I have a choice between using RAID 1 SCSI or, having a twin disk with a periodic backup from one disk to another - something like ultrabac. I'm planning to avoid tape backups altogether and rely on disk based backups for speed and cost reasons.

I appreciate that RAID 1 has the benefit of hot standby should the harddrive fail. But, if somebody does some file damage to the main RAID disk, it gets copied onto the backup disk.

With the ultabac solution, there is more downtime should a failure occur with the possibility of lost data ?

Are these two statements correct ?

In either of these situations, should the server hardware itself fail, can I (in principle) rip the disk out of the server and pop it into another server with a few config changes e.g. a job that could be done within a few hours downtime?

Which would you do in my situation. We are talking about a valuable webserver with a SQL Server backend.


Please help !

Thanks

 

 

 

 

Top